Headset Voice Extraction Using Internal and External Microphones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional headsets face challenges in reliably detecting and extracting a user's own voice amidst background noise, leading to suboptimal performance in speech recognition systems due to bulky designs, limited noise rejection, and noise amplification issues with existing microphone configurations.
Innovation Solution
A headset system employing both internal and external microphones, where noise reduction is achieved by filtering the internal microphone signal to generate a noise-reduced signal indicative of the user's voice content, using a transfer function that inversely represents the filtering through the earpiece, and further processing includes equalization and residual noise reduction to enhance voice quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If near field microphone array techniques and external microphones are used for noise cancellation, then noise rejection capability is improved, but headset design becomes bulky and prone to physical damage
Solution Approach 1:
The patent combines internal and external microphone signals through adaptive filtering to achieve noise cancellation. The external microphone captures ambient noise, while the internal microphone captures the user's voice. An adaptive filter processes these signals to separate voice from noise, merging the functionality of multiple microphones into a unified noise cancellation system without requiring additional bulky components.
Solution Approach 2:
The headset microphones serve multiple functions: the external microphone captures both noise and voice, while the internal microphone captures voice and bone conduction signals. These same microphones are used for both noise cancellation and voice enhancement, eliminating the need for separate dedicated microphones for each function.
2Object-affected harmful factors
If beamforming techniques with multiple external microphones are used, then noise rejection is improved, but the limited space on headset restricts microphone array size and directivity
Solution Approach 1:
The patent transitions from relying solely on spatial arrangement of multiple external microphones (2D plane) to utilizing the third dimension of bone conduction through the skull. The internal microphone captures bone conduction signals that bypass external noise, adding a new dimensional approach to noise rejection that isn't constrained by headset space.
3Measurement precision
If conventional noise reduction is applied to external microphone signal for equalization, then high frequency loss is compensated, but significant noise amplification occurs
Solution Approach 1:
The patent extracts the noise component from the external microphone signal using adaptive filtering before applying equalization. By separating the noise from the voice signal beforehand, the subsequent equalization process can boost high frequencies without simultaneously amplifying the noise, thus achieving frequency response correction without the harmful side effect of noise amplification.
Solution Approach 2:
The signal processing is segmented into distinct stages: first, adaptive filtering separates noise from voice in the external microphone signal; second, equalization is applied to the filtered signal to correct frequency response. This segmentation allows each processing stage to focus on its specific task without compounding harmful effects.
4Object-affected harmful factors
If simple suppression based noise reduction is used, then stationary background noise is reduced, but other noise such as competing talker noise is not effectively handled
Solution Approach 1:
The patent employs adaptive filtering that dynamically adjusts its parameters based on the characteristics of the input signals. The adaptive filter continuously monitors the external and internal microphone signals and modifies its filtering behavior in real-time, enabling it to handle different types of noise (stationary background noise, competing talkers, dynamic environments) rather than being limited to fixed suppression of stationary noise only.
Data Source
AI summary
Methods and systems employing an internal microphone and an external microphone of a headset to capture own voice content in the presence of noise, extract the own voice content from background noise (by performing noise reduction on the microphone outputs to generate a noise reduced signal indicative of the own voice content), and optionally also perform voice activity detection to identify segments of own voice presence or absence. In some embodiments, the external microphone is employed to capture the own voice content, the internal microphone signal is employed to infer the noise captured by the external microphone, and the inferred noise is subtracted from the external microphone signal to generate the noise reduced signal. Aspects include methods performed by any embodiment of the system, and a system or device configured (e.g., programmed) to perform any embodiment of the method.